Modeling of user behavior has been applied to content consumption as it relates to the behavior of users consuming the content. Current behavioral modeling focuses on predicting user preferences for what type of content the user prefers. Thus, behavioral modeling can determine that a user is a sports fan and can therefore be used to recommend a sporting event to the user. However, users consuming content exhibit different consumption behaviors based on genre of content consumed. For example, users consuming content in the sports genre exhibit different behaviors with regard to how content is consumed from users consuming content in the movies genre. Current behavioral modeling fails to predict how users consume content, but rather focus on what content the users might want to consume. These and other shortcomings are addressed by the approaches set forth herein.